Some observers believe that there will be such growth in the travelling public and such constraint on airport and airspace usage that, by early in the 2000s. very large passenger carrying aircraft will be required. By ‘very large’ is meant anything from 1,000 to 2,000 seaters. Whilst the problems which come to mind when musing on this possibility are mostly those of ‘passenger flow’ in the widest sense and rather less related to the vehicle itself, there are some issues related to the vehicle which suggest themselves. One is the choice of fuselage cross section.

Because cabins have to be pressurised (10 psi is a typical figure for cruise altitudes up to say 40,000 ft), it has been the norm to utilise circular cross sections (or double-bubble type) since the pressure forces are contained by hoop loads only, without frame bending, which affords a light and fatigue-insensitive structure (Fig. 1).

The procedures employed for the design of a closed-circuit, boundary layer wind tunnel are described. The tunnel was designed for the generation of relatively large-scale, two-dimensional boundary layers with Reynolds numbers, pressure gradients and free-stream turbulence levels typical of the turbomachinery environment. The results of a series of tests to evaluate the tunnel performance are also described. The flow in the test section is shown to be highly uniform and steady, with very low (natural) free-stream turbulence intensities. Measured boundary layer mean and fluctuating velocity profiles were found to be in good agreement with classical correlations. Test-section free-stream turbulence intensities are presented for grid-generated turbulence: agreement with expectation is again found to be good. Immediate applications to the tunnel include friction drag reduction and boundary layer transition studies, with future possibilities including flow separation and other complex flows typical of those found in gas turbines.

Combined effects of geometrical and material non-linearities on a cantilever beam having symmetrical cross-section about its central axis with a rotational distributed loading are studied. It is assumed that the stress-strain relation in compression is identical to that in tension. Due to this, the neutral axis coincides with the central axis of the beam. The problem is formulated by means of an integral equation which is suitably converted to a system of nonlinear ordinary differential equations which are solved using a simple and accurate numerical method. Details of the load deflection characteristics for an I-beam and for a beam of rectangular cross-section are presented.

A three-dimensional integral formulation for the solution of the full potential equation and the associated numerical algorithm, the field-integral method, are presented. The method is a direct extension of a two-dimensional method and in particular retains the simple grid generation requirements noted in that method. Results are presented for the flow over body shapes and a complex winglet configuration, and are compared with existing transonic methods and experiments with good agreement. The further work necessary to provide a fast, robust method for use in design is outlined.